试试sudo pip install numpy——upgrade——ignore-installed。

它适用于Mac OS 10.11。

如果上面的shell不能工作，你应该关闭“Rootless”。

我也遇到过同样的问题，下面是我的解决方法。

我尝试了pip install -U numpy，但它没有升级numpy，但conda install为我工作

ImportError: numpy.core.multiarray failed to import
admin@MacBook-Air$ pip install -U numpy
Requirement already up-to-date: numpy in /Users/admin/anaconda/lib/python2.7/site-packages
admin@MacBook-Air$ python
Python 2.7.12 |Anaconda 2.4.0 (x86_64)| (default, Jul  2 2016, 17:43:17) 
[GCC 4.2.1 (Based on Apple Inc. build 5658) (LLVM build 2336.11.00)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
Anaconda is brought to you by Continuum Analytics.
Please check out: http://continuum.io/thanks and https://anaconda.org
>>> import numpy
numpy.version.version
>>> numpy.version.version
'1.7.1'
>>> exit
Use exit() or Ctrl-D (i.e. EOF) to exit
>>> 
admin@MacBook-Air$ 
admin@MacBook-Air$ conda install numpy
Fetching package metadata .......
Solving package specifications: ..........

Package plan for installation in environment /Users/admin/anaconda:

The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    scikit-learn-0.18.1        |      np111py27_0         4.9 MB

The following packages will be UPDATED:

    numexpr:      2.3.0-np17py27_0  --> 2.6.1-np111py27_1 
    numpy:        1.7.1-py27_2      --> 1.11.2-py27_0     
    scikit-learn: 0.14.1-np17py27_1 --> 0.18.1-np111py27_0
    scipy:        0.13.2-np17py27_1 --> 0.18.1-np111py27_0

Proceed ([y]/n)? y

Fetching packages ...
scikit-learn-0 100% |#################################################################| Time: 0:00:16 312.60 kB/s
Extracting packages ...
[      COMPLETE      ]|####################################################################################| 100%
Unlinking packages ...
[      COMPLETE      ]|####################################################################################| 100%
Linking packages ...
[      COMPLETE      ]|####################################################################################| 100%

如果你卸载并重新安装或安装了更新，但它仍然不能工作，特别是如果你用cx_freeze打包:

尝试在导入cv2之前先导入numpy来更改代码。

所有这些不同的答案都表明出现了许多不同的问题，例如ImportError: numpy.core.multiarray导入失败。

==>所以在堆栈回溯之前查找更多的错误信息

例如，重现一个这样的问题的步骤(这些步骤在numpy 1.20发布之前都很好):

# set up a virtualenv
pyenv virtualenv 3.8.7 pip-issue-9542
pyenv local pip-issue-9542

# update pip, setuptools
pip install -U pip setuptools

# install 2 pips
pip install --no-cache-dir numpy==1.19.5 cvxpy==1.1.3

# try to load cvxpy
python -c 'import cvxpy'

输出:

RuntimeError: module compiled against API version 0xe but this version of numpy is 0xd
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/usr/local/var/pyenv/versions/pip-issue-9542/lib/python3.8/site-packages/cvxpy/__init__.py", line 18, in <module>
    from cvxpy.atoms import *
  File "/usr/local/var/pyenv/versions/pip-issue-9542/lib/python3.8/site-packages/cvxpy/atoms/__init__.py", line 20, in <module>
    from cvxpy.atoms.geo_mean import geo_mean
  File "/usr/local/var/pyenv/versions/pip-issue-9542/lib/python3.8/site-packages/cvxpy/atoms/geo_mean.py", line 20, in <module>
    from cvxpy.utilities.power_tools import (fracify, decompose, approx_error, lower_bound,
  File "/usr/local/var/pyenv/versions/pip-issue-9542/lib/python3.8/site-packages/cvxpy/utilities/power_tools.py", line 18, in <module>
    from cvxpy.atoms.affine.reshape import reshape
  File "/usr/local/var/pyenv/versions/pip-issue-9542/lib/python3.8/site-packages/cvxpy/atoms/affine/reshape.py", line 18, in <module>
    from cvxpy.atoms.affine.hstack import hstack
  File "/usr/local/var/pyenv/versions/pip-issue-9542/lib/python3.8/site-packages/cvxpy/atoms/affine/hstack.py", line 18, in <module>
    from cvxpy.atoms.affine.affine_atom import AffAtom
  File "/usr/local/var/pyenv/versions/pip-issue-9542/lib/python3.8/site-packages/cvxpy/atoms/affine/affine_atom.py", line 22, in <module>
    from cvxpy.cvxcore.python import canonInterface
  File "/usr/local/var/pyenv/versions/pip-issue-9542/lib/python3.8/site-packages/cvxpy/cvxcore/python/__init__.py", line 3, in <module>
    import _cvxcore
ImportError: numpy.core.multiarray failed to import

重要的线索是RuntimeError:模块根据API版本0xe编译，但numpy的这个版本是0xd。

搜索它可以让你找到pip issue 9542和cvxpy issue 1229，它们解释了哪里出了问题。

本例中的变通方法是(a)更新为cvxpy>=1.1.10，或(b)更新为numpy>=1.20，或(c)两者都更新。您必须处理库更改，因为如果不深入pip的“无构建隔离”世界，原始环境是不可复制的。

在试图导入Pytorch时遇到这种情况。解决方法是反复卸载或删除numpy，直到我的Ubuntu上没有任何版本，然后安装最新版本。在我的情况下，pip遇到访问被拒绝的权限错误，可能是因为与conda冲突。

执行sudo pip uninstall numpy命令。如果它拒绝从<location>中删除包(确保它是特定包的位置，而不是所有包-这是不可逆的)，运行sudo rm -r <location>。重复以上步骤，直到第一个命令不起任何作用，并且import numpy无法在Python shell中找到包。

使用pip安装numpy。

对我来说，这是两部分。第一:

pip uninstall numpy

然后@Oli Blum回答(谢谢):

conda install -c conda-forge numpy

这招对我很管用